Marc Brisson Full Professor Université Laval, Canada IMPACT OF HPV IMMUNIZATION STRATEGIES & POTENTIAL FOR CERVICAL CANCER ELIMINATION SAGE meeting October 24, 2018 Geneva 1
Marc BrissonFull Professor
Université Laval, Canada
IMPACT OF HPV IMMUNIZATION STRATEGIES & POTENTIAL FOR CERVICAL CANCER
ELIMINATION
SAGE meetingOctober 24, 2018 Geneva
1
Questions
A. What is the potential for Cervical Cancer elimination with HPV immunization?
B. What is the population-level effectiveness and cost-effectiveness of different HPV immunization schedules and strategies ?
2
POTENTIAL FOR CERVICAL CANCER ELIMINATION :
A COMPARATIVE MODELING STUDY
M Brisson, J Kim, K Canfell,C Gopalappa, E Burger, D Martin, E Benard, Kate Simms, M Drolet, S Sy, C Regan, A Keane, M Smith, C Pretorius,
MC Boily, N Broutet & R Hutubessy
Université Laval, Canada; Harvard T.H. Chan School of Public Health, USA; Cancer Council NSW, Australia; University of Massachusetts Amherst, USA; CHU de Québec-Université Laval Research Center, Canada; Avenir Health, USA; Imperial
College, UK; World Health Organization, Switzerland
SAGE meetingOctober 24, 2018 Geneva
3
Call for action to eliminate cervical cancer
• In May, the WHO Director-General made a global call for action to eliminate cervical cancer as a public health problem&
• Elimination of cervical cancer as a public health problem is different to elimination of an infectious disease
– Not reduction to 0 incidence
– Control of cervical cancer at a low disease incidence
– Requires clear well defined threshold
– Previously used for other diseases by WHO
• Neonatal Tetanus (NT): 1 NT per 1000 live births per yr
• Congenital syphilis : Case rate of ≤50 per 100 000 live births
&: WHO Director-General Speech - http://www.who.int/reproductivehealth/DG_Call-to-Action.pdf4
Call for action to eliminate cervical cancer Key questions that must be addressed
• What is the definition of cervical cancer elimination as a public health problem?
– What outcome? Cancer incidence? Mortality? % Reduction?
– Same for every country?
– Pragmatic? Optimistic?
• What combination of screening and vaccination strategies can lead to elimination (for different definitions)?
• When could elimination be reached, for different strategies and countries?
• What is the most efficient/cost-effective strategy to reach elimination?
5
Need for mathematical models
• Mathematical models provide a formal framework to examine key elimination questions– project long-term population-level effects (e.g., herd immunity)
– evaluate multiple strategies under varying assumptions
• However, models require many simplifications & assumptions which leads to uncertainty in the validity of predictions– can create uncertainty for decision makers
• WHO initiated a model comparison to help provide guidance for cervical cancer elimination – the Cervical cancer elimination modeling consortium was created
6
• Model Selection– Dynamic model– Model includes vaccination & screening– Independent model that has been peer reviewed/published
• Step 1: Understand Model differences/ Potential for elimination– Use selected models from Step 0– Examine 14 simplified screening & vaccination scenarios– Selection of calibrated countries – Compare model structures & results
• Step 2: Examine key questions for a selected set of countries– Use the selected models– Examine 26 realistic screening & vaccination scenarios– Model a wider range of countries
• Step 3: Estimate global model predictions– Use 3 of 4 selected models – Examine 3 screening & vaccination scenarios– Estimate global predictions
• Step 4: Economic analysis – Use 3 of 4 selected models – Examine the cost & cost-effectiveness of elimination
Systematic comparative modeling approach
MARCH
• Policy 1 Model– Lead: Karen Canfell– Team: Kate Simms, Adam Keane, Megan Smith– Institution: Cancer Council NSW, Australia
• Harvard Model– Lead: Jane Kim– Team: Emily Burger, Stephen Sy, Catherine Regan– Institution: Harvard, USA
• HPV-ADVISE Model– Lead: Marc Brisson– Team: Mélanie Drolet, JF Laprise, Dave Martin, Élodie Bénard, Guillaume Gingras, Iacopo
Baussano, Marie-Claude Boily, Mark Jit – Institution: U Laval, Canada; Imperial College, UK; LSHTM, UK; IARC, France
• Spectrum Model– Leads: Chaitra Gopalappa & Carel Pretorius– Institution: U Massachusetts & Avenir Health, USA
8
Global predictions
78 Low & Lower Middle Income Countries
2 vaccination/screening scenarios
9
• S1 - Scenario 1: – Girls-only vaccination (90% coverage, 9-14 yr old)
– No change in Screening
• S2 - Scenario 2:
– Girls-only vaccination (90% coverage, 9-14 yr old)
– 2 lifetime screens at 35 and 45 yrs old– High Screening ramp-up (45%, 70%, 90% in 2023, 2030, 2045, respectively)
• All scenarios:– Screening: HPV testing, 100% treatment efficacy, 10% Lost to follow-up– Vaccine: Lifelong duration, 100% efficacy, HPV16/18/31/33/45/52/58
Vaccination & Screening scenarios
10
Years
Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58
No further changeVaccination only (S1)
Cerv
ical
can
cer
inci
dence
(per
100,0
00)
Dynamics of elimination Consistency in model predictionsHigh vaccination coverage & Screening Ramp-up
Low income countries Lower middle income countries
HPV-ADVISE
Harvard
Policy1
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
HPV-ADVISE
Harvard
Policy1
11
Years&. Mean predictions; Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58, HPV testing
Vaccination, 2 lifetime screens (S2)
Vaccination only (S1)
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
Time to
<10/100,000 <4/100,000
LIC LMIC LIC LMIC
S1 2078 2064 X 2081
S2 2066 2050 2083 2069
Cerv
ical
can
cer
inci
dence
(per
100,0
00)
Dynamics of elimination Impact over time Low (LIC) & Lower Middle Income Countries (LMIC), High vaccination coverage & Screening Ramp-up
12
YearsMean prediction of models; Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58, HPV testing
Vaccination, 2 lifetime screens (S2)
Vaccination only (S1)
Cerv
ical
can
cer
inci
denc
e (p
er 1
00,0
00)
Dynamics of elimination Impact over time – under 45 year oldsLow (LIC) & Lower Middle Income Countries (LMIC), High vaccination coverage & Screening Ramp-up
0
1
2
3
4
5
6
7
8
9
10
2020 2040 2060 2080 2100 2120
0
1
2
3
4
5
6
7
8
9
10
2020 2040 2060 2080 2100 2120
LICLMIC<45yrsNo further change
13
Years&. Mean prediction over 100 years; Adjusted cases averted for 2015 population; Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58, HPV testing
No further change
Vaccination, 2 lifetime screensVaccination only
Cerv
ical
can
cer
inci
dence
(per
100,0
00)
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
Dynamics of elimination Incremental benefits of strategiesHigh vaccination coverage & Screening Ramp-up
Cases averted&
(vs no change)Incremental
cases avertedLIC LMIC LIC LMIC
3.5 M 11.5 M 3.5 M 11.5 M4.3 M 14.2 M 0.8 M 2.7 M
Low income countries Lower middle income countries
14
Country specific predictionsLow income & Lower Middle income countriesVaccination only (S1)
&. Median prediction; Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58 15
Country specific predictionsLow income & Lower Middle income countries by regionVaccination only (S1)
&. Median prediction; Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58
Time of elimination in all countries in region
(lower & upper threshold)
26
Country specific predictionsLow income & Lower Middle income countriesVaccination, 2 lifetime screens (S2)
&. Median prediction; Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58, HPV test 27
&. Median prediction; Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58, HPV test
Country specific predictionsLow income & Lower Middle income countries by regionVaccination, 2 lifetime screens (S2)
Time of elimination in all countries in region
(lower & upper threshold)
38
0
2
4
6
8
10
12
14
0 20 40 60 80 100 120
Country specific elimination predictionsImpact of starting cervical cancer (CC) incidenceLow & Lower middle income countries, Vaccination only (S1)
&. Harvard & HPV-ADVISE; High vaccination coverage and screening ramp-up, Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58, HPV test
Cervical cancer incidence 2020 (per 100,000)
Cerv
ical
can
cer
inci
denc
e 21
20(p
er 1
00,0
00)
40
0
20
40
60
80
100
120
0 20 40 60 80 100 120
Country specific predictions Absolute Reductions (AR) in Cervical cancer (CC) & EliminationLow & Lower middle income countries, Vaccination only (S1)
&. Harvard & HPV-ADVISE; High vaccination coverage and screening ramp-up, Girls-only vaccination, Vaccine protection=HPV16/18/31/33/45/52/58, HPV test
AR=90/100,000
AR=18/100,000
41
Cervical cancer incidence 2020 (per 100,000)
Cerv
ical
can
cer
inci
denc
e 21
20(p
er 1
00,0
00)
SummaryGlobal analysis: 3 optimistic vaccination & screening coverage scenarios
What strategies lead to elimination?
• Girls-only vaccination leads to incidence <10/100,000 w-yrs without screening in most countries/regions– <15/100,000 w-yrs in Sub-Saharan Africa
• Girls-only vaccination & 2 lifetime screens leads to incidence <4/100,000 w-yrs in most countries/regions– <10/100,000 w-yrs in Sub-Saharan Africa
When does elimination occur? • Average within LIC/LMIC: 2045-2060
– UMIC/HIC: elimination occurs earlier
• 100% of countries: 2085-2105 • Depends on the strategy & threshold
&. Low (LIC), Lower Middle Income (LMIC), Upper Middle Income (LMIC), High Income Countries (HIC), Vaccine protection=HPV16/18/31/33/45/52/58, HPV test 42
Sensitivity Analysis
What is the impact of:
Vaccination Strategies & Coverage? Number of Screens?
Vaccine characteristics?
43
Years
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
No further change
80% Girls-only vaccination90% Girls-only vaccination
80% Girls&Boys vaccination80% Girls&Boys vaccination, Catch-up
Cerv
ical
can
cer
inci
dence
(per
100,0
00)
VIETNAM UGANDA
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
Time to<10/100,000 <4/100,000VN UG VN UG
2066 2084 2086 X2068 X X X2064 2084 2084 X2058 2077 2077 X
Mean predictions; Vietnam: Policy-1/HPV-ADVISE; Uganda: Harvard/HPV-ADVISE; Vaccine duration=lifelong; VE=100%
Dynamics of elimination Impact of vaccinationNo change in screening
46
Years
Cerv
ical
can
cer
inci
dence
(per
100,0
00)
VIETNAM UGANDA
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
No further change
Efficacy against HPV16/18 onlyEfficacy against HPV16/18/31/33/45/52/58
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
Dynamics of elimination Impact of number of HPV types protected 80% Girls-only vaccination, high screening ramp-up, 2 screens
Mean predictions; Vietnam: Policy-1/HPV-ADVISE; Uganda: Harvard/HPV-ADVISE; Vaccine duration=lifelong; VE=100%
Time to<10/100,000 <4/100,000VN UG VN UG
2054 2073 2077 20972054 2082 2089 X
48
Years
No further change
Vaccine Duration = 20 yearsVaccine Duration = lifelong
Cerv
ical
can
cer
inci
dence
(per
100,0
00)
VIETNAM UGANDA
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
0
10
20
30
40
50
60
70
2020 2040 2060 2080 2100 2120
Dynamics of elimination Impact of duration of vaccine protected 80% Girls-only vaccination, high screening ramp-up, 2 screens
Mean predictions; Vietnam: Policy-1/HPV-ADVISE; Uganda: Harvard/HPV-ADVISE; Vaccine duration=lifelong; VE=100%
Time to<10/100,000 <4/100,000VN UG VN UG
2054 2073 2077 20972054 X X X
49
Sensitivity analysis40 scenarios, 12 countries
• Greatest additional benefits (cancers cases averted over time):– Vaccination of Girls-only with high coverage (vs no vaccination)
– 2 lifetime screens (vs current screening)
– Multi-cohort vaccination (vs single-cohort vaccination) – No impact on elimination
• Screening or multi-cohort vaccination accelerates elimination (5-15 yrs)
• Smallest additional impact– Vaccinating boys (vs Girls-only) if Girls-only coverage is high
• Long-term duration of vaccine protection is required for elimination
50
SUMMARY• Models produced consistent findings
• Countries with cervical cancer incidence < 30/100,000 w-yrs– >80% Girls-only vaccination coverage could lead to elimination
without changes to screening
• Countries with cervical cancer incidence ≥ 30/100,000 w-yrs– elimination is highly dependent on the threshold used
– high screening & vaccination coverage, and a broad spectrum vaccine is required
– hardest to eliminate but have greatest absolute reductions in incidence
– countries with incidence ≥ 70/100,000 w-yrs may not reach elimination
• Long-term vaccine protection is needed (>20 years)– particularly for higher cervical cancer incidence countries
51
SUMMARY• Greatest additional benefits:
– Vaccination of Girls-only (vs no vaccination)
– 2 lifetime screens (vs no screening)
– Multi-cohort vaccination (vs single-cohort vaccination)
– Screening at least once in a lifetime & multi-cohort vaccination accelerates elimination by about 10 years
• Results are most sensitive to:– Definition of elimination - Threshold
– Starting cervical cancer incidence
• Future work:– Examine the cost and cost-effectiveness of elimination
– Identify the most efficient strategies for elimination
52
Brisson M1,2, Jit M3,4, Bénard É2, Martin D2, Drolet M2, Laprise JF2, Boily MC5, Alary M1,2, Baussano I6, Gingras G2,
Pérez N2, Hutubessy R7
1. Université Laval, 2. Centre de recherche du CHU de Québec, 3. London School of Hygiene & Tropical Medicine, 4. Public Health England, 5. Imperial College,
6. International Agency for Research on Cancer (IARC), 7. World Health Organization (WHO)
IMPACT OF DIFFERENT HPV IMMUNIZATION SCHEDULES AND STRATEGIES
SAGE meetingOctober 24, 2018 Geneva
2853
Objective• Examine the population-level effectiveness and cost-effectiveness of
HPV immunization of different schedules and strategies in Low and Lower Middle Income Countries, using:
– Predictions from Mathematical Models
• Girls-only HPV immunization (HPV2 or HPV4 vs HPV9)
• Gender-neutral HPV immunization (vs Girls-only)
• Multiple age cohort HPV immunization (vs single age cohort)
Schedules/strategies
54
REF: 1. Brisson, JNCI 2015; &: Demographic and Health Surveys, Multiple Indicator Survey, ICO information Centre on HPV and Cancer, United Nations Statistics Division, HIV and AIDS HUB for Asia Pacific-Evidence to action, WHO Global Health Observatory data repository, literature reviews, and original studies from IARC and Dr. M Alary (see back-up slides for references & model fit)
MethodsModeling - Population-level effectiveness & herd effects
HPV-ADVISE (Agent-based Dynamic model for VaccInation & Screening Evaluation)1
• Transmission-dynamic model of HPV infection and disease (includes herd immunity)
• Models 18 HPV types:
– Types included in the 9-valent vaccine (HPV-6/11/16/18/31/33/45/52/58)
– 9 other high risk types
• Fit HPV-ADVISE to Canada, India, Vietnam, Benin, Nigeria and Uganda&
– Demographic and sexual behaviour
– HPV prevalence and cervical cancer incidence (age and type-specific)
– Data from international databases and original studies&
55
Question: Girls-only immunization • What is the incremental effectiveness and cost-effectiveness for cervical
cancer prevention of different HPV vaccines based on Girls-only immunization?
• Girls-only HPV vaccination (vs no vaccination)
• High population-level effectiveness & strong herd effects
• Highly cost-effective, irrespective of vaccine used
• Main driver: Prevention of HPV-16/18 related cervical cancer
• Cost-effective even when excluding herd immunity, cross-protection & benefit from reducing non-cervical diseases
• HPV9 vaccine Girls-only vaccination (vs HPV2 or HPV4) • Likely cost-effective (vs HPV2 or HPV4) in HIC & LMIC unless
• very strong cross-protection from HPV2 or HPV4 is expected
• HPV9 priced too high
• Main drivers: Cross-protection from HPV2/4 and vaccine price
Key modeling results
REF: HPV-ADVISE, SAGE WG Meeting 2018; Ng, Vaccine 2018; Jit, Lancet Global Health 2015; Fesenfeld, Vaccine 2013 56
Question: Gender-neutral immunization• What is the incremental effectiveness and cost-effectiveness of adolescent
Gender-neutral HPV immunization compared to Girls-only HPV immunization?
Incremental effectiveness• HIC: Small additional benefits of vaccinating boys (even at low coverage )
• LMIC: Greater added benefit of vaccinating boys than in HIC
• HIC & LMIC: Increasing coverage in girls provides greater impact than including boys
Cost-effectiveness of vaccinating girls & boys (vs girls-only)
• HIC: Unlikely cost-effective IF vaccine coverage is high in girls
• LMIC: May be cost-effective even if coverage in girls is high
• LMIC: More cost-effective to increase coverage in girls when coverage is low
Main drivers
• Magnitude of herd effects by Girls-only vaccination / Burden of anogenital warts and HPV-related cancers
Key modeling results
57REF: HPV-ADVISE, SAGE WG Meeting 2018; Ng, Vaccine 2018; Brisson, Lancet Public Health
Question• What is the incremental effectiveness and cost-effectiveness for cervical
cancer prevention of different HPV vaccines based on Girls-only immunization?
• Girls-only HPV vaccination (vs no vaccination)
• High population-level effectiveness & strong herd effects
• Highly cost-effective, irrespective of vaccine used
• Main driver: Prevention of HPV-16/18 related cervical cancer
• Cost-effective even when excluding herd immunity, cross-protection & benefit from reducing non-cervical diseases
• HPV9 vaccine Girls-only vaccination (vs HPV2 or HPV4) • Likely cost-effective (vs HPV2 or HPV4) in HIC & LMIC unless
• very strong cross-protection from HPV2 or HPV4 is expected
• HPV9 priced too high
• Main drivers: Cross-protection from HPV2/4 and vaccine price
Key modeling results
REF: HPV-ADVISE, SAGE WG Meeting 2018; Ng, Vaccine 2018; Jit, Lancet Global Health 2015; Fesenfeld, Vaccine 2013 58
Thank you!